existence of a low-level wind maximum during periods of 

 temperature inversions (Thuiilier and Lappe 1964). 



Windspeed observations on high mountain slopes at 

 night are not common. Nevertheless, Lee' noted ridgetop 

 windspeeds were considerably stronger at 0300 P.s.t. than 

 at 0800 P.s.t. in the Blue Mountains of northeastern 

 Oregon. Wind profile measurements at Burns, Oreg., 

 indicated a definite wind maximum at low levels during the 

 night. Perhaps Lee was the first to refer to Blackadar's 

 theory as an explanation of the observed strong mountain 

 winds. And as noted before. Rider (1966) also observed 

 low-level jet winds over mountainous terrain. 



The recorded occurrence of low-level windspeed 

 maxima in conjunction with a forest fire in mountain 

 areas is apparently quite rare. Byram (1954) observed jet 

 currents over fires in South Dakota and California. Small 

 (1957) reports that careful examination of the Boise 

 Weather Bureau records indicates the presence of a low- 

 level jet point on each day of significant spread by the 

 Robie Creek Fire of 1955 in southern Idaho. Also, Finklin 

 (1973) discussed the possibility of a low-level jet wind 

 affecting the major run of the Sundance Fire in northern 

 Idaho in the fall of 1967. 



SUNDANCE FIRE - LOW-LEVEL 

 JET 



There is strong evidence that a low-level jet wind 

 also occurred 3 days before the major run of the Sundance 

 Fire. An account -^of this earlier outbreak of the fire is 

 given by Anderson (1968). 

 Late in the evening of August 29, at 2220, Priest Lake 

 Timber Protective Association Headquarters near 

 Coolin, Idaho, received word the fire had jumped the 

 line and was out of control. Men and equipment were 

 evacuated to headquarters. The fire was observed 

 rolling down the hill in the Lee Cl-eek drainage, a 

 northeast wind had prevailed throughout the day and it 

 is assumed this wind, coupled with the normal nighttime 

 downslope currents, resulted in a wind-driven fire 

 moving downslope. Information available from 

 observers indicated the winds were 20-25 mi/h on the 

 fire, but there was calm at Cavanaugh Bay near Coolin. 

 The ground and crown fires were observed to move as a 

 single front, with spotting up to one-half mile ahead. 

 The main advance took place between 2230, August 29, 

 and 0200, August 30, which would give an average rate of 

 spread of 0.80 mi/h. 



There was no direct measurement of a surface inversion 

 in the area, but there is little doubt that one occurred 

 during the night of August 29, because the region was 

 under the influence of a high-pressure ridge at the time 

 (Finklin 1973). Also, in an unpublished report (G. A. 

 Verdall, letter in files at the Northern Forest Fire 

 Laboratory, dated Oct. 5, 1967), it was reported, "Cooler 

 air had filled the valley bottom; therefore there was no wind 

 at our location (Cavanaugh Bay)." Verdal went on to say, 

 "What made the fire behavior seem unusual to me was that I 

 was standing in calm air watching a wind driven fire." Also 

 he said, "To me, the fire behavior witnessed at that time was 

 more spectacular and more unexpected than the major run 

 the fire made on September 1." 



The fire burned from the top of Sundance Mountain 

 (6,000 ft [1 830 m] m.s.l.) down to about the 4,000 ft 

 (1 220 m) level (fig. 2). The windspeed wa^ quite strong 

 on the ridgetop while it was calm some 2 miles away at 

 Cavanaugh Bay (2,500 ft [762 m] m.s.i.). According to 

 the low-level jet-wind theory, this is not an unexpected 

 result, because the ridge was exposed to the jet wind while 

 Cavanaugh Bay was protected by the temperature 

 inversion layer. Upper air information was not available 

 for the fire site, but it was available for Spokane, Wash., 

 about 50 miles (80 km) southwest of the fire. The data 

 clearly show the presence of a low-level jet wind during 

 this early fire outbreak (fig. 3). 



Figure 2.— Sundance Mountain after the 

 fire's night run on August 29, 1967. 



The windspeed on the night of August 29-30 was from 

 the east to northeast reaching 15 to 20 mi/h (6.7 to 8.9 m/s) 

 at a nearby mountaintop fire-weather station (Lunch Peak). 

 "These winds were associated with a cool surface high to 

 the northeast, moving southeastward across Alberta and 

 Saskatchewan" (Finklin 1973). The surface map of 

 0600Z, August 30 (2200 P.s.t., August 29, time of the fire 

 outbreak) revealed the presence of the cold high-pressure 

 system mentioned by Finklin and a small, warm low- 

 pressure system to the southwest of the fire site over 

 eastern Washington. This was just the type of synoptic 

 situation required to produce an enhanced low-level 

 jet wind; that is, opposing thermal and geostrophic wind 

 components. The windspeed profile at 1600 on August 

 29 (fig. 3) was nearly constant with height up to 8,000 

 ft (2 440 m). At 2200 (August 29), the time of the fire 

 outbreak, the jet was well formed, reaching a velocity of 

 20 knots (east wind at 23 mi/h [10 m/sj), at an elevation of 

 5,000 ft (1 525 m). The jet was still present early 

 the next morning as shown by the 0400 sounding of 

 August 30; still east at 20 knots (10 m/s). Strong winds 

 continued at Lunch Peak during the morning hours 

 decreasing to 10 to 12 mi/h (4.5 to 5.4 m/s) at 1400. The 

 Spokane data for 1100 (August 30) is missing, but the 



4 



